168 PROF. n. G. SEELEY ON SAURODESMUS ROBKRTSONI. 



radial side of the bone above the articulation to a sharp edge, which 

 makes the longitudinal contour convex on that side at the distal end, 

 and in the dissimilar form of the distal articular surface with its well 

 ossified detail and definition. Even disregarding the absence from 

 the fossil of a radial groove or perforation, the characters enumerated 

 seem to me to outweigh the slight resemblances to Chelonians which 

 it wseems to present, and it would follow that there is practically no 

 evidence of value in favour of the Chelonian nature of the fossil. 



At first sight the bone (figs. 1-6) is not very like the humerus of a 

 Crocodile, yet its afiinities with the Crocodilia are more important. 

 Careful consideration of the radial process in the fossil shows that 

 its inner prolongation is a sharp almost knife-edge, as in Crocodiles, 

 and that the proximal articular part of the bone prolonged the shaft 

 beyond the radial process in a ^Yixy only paralleled among Crocodiles. 

 On the other hand, the radial crest is never reflexed forward so much 

 in existing Crocodiles, and there is no Crocodile in which the ulnar 

 border is compressed to a sharp muscular edge, though there is, 

 perhaps, a faint suggestion of such a ridge in a proximal angle of 

 the ulnar tuberosity ; and in Gavials, some recent species of Croco- 

 dilus, and some fossil Crocodiles like Crocodilus Haatingioi, the 

 character is slightly marked. 



The bone as a whole is much more expanded at both ends than 

 in Crocodiles, and is even straighter, but the distal articulation is 

 essentially the same in plan, with like details of cond^^lar structure 

 and a like compression of the bone on the radial side, though the 

 ridge in the living types is very slight compared \Aith its development 

 in this fossil. These are, however, homologous characters, and 

 Crocodiles have the limb-bones hollow ; so that, as the indi- 

 cations from the proximal and distal ends and from the internal 

 structure all point to the same result, it may be concluded that the 

 Linksfield fossil indicated a primitive Crocodilian stock ; and that 

 the intensified characters which it shows are feebly preserved in its 

 surviving representatives. The chief differences from Crocodiles are 

 that the radial crest is directed more forward and less downward ; 

 that the ulnar side is sharply compressed, ends in a muscular ridge, 

 and has a convex curve ; that the shaft is straighter ; and that the 

 distal end is relatively wider, with its radial border much more 

 compressed. 



In Lizards there are some approximations in these points which 

 are worthy of remark. The extremities of the bone are more ex- 

 panded transversely than among Crocodiles, but then the shaft is 

 twisted. Both the radial border at the distal end and the ulnar 

 border towards the proximal end are compressed in Lizards, but 

 then the distal end has enormous articular condyles equally unlike 

 those of the Linksfield fossil and Crocodiles. The concavity below 

 the head of the humerus is more open transversely in Lizards, but 

 then a strong rounded ridge connects the radial crest with the 

 articular head. In no respect, however, either in characters of the 

 proximal or distal end, can Lizards be said to approach so near to 

 the fossil as do Crocodiles. Nevertheless, there may be a tendency 



